RP2350A + ADS1220 Molecule

@@ -0,0 +1,53 @@+# RP2350A + ADS1220 Molecule — Function++24-bit thrust acquisition. RP2350A (QFN-60) + ADS1220 + W25Q16JV + AP2112K-3.3, two differential+load-cell channels. 26.40 × 36.00 mm, 4-layer, 70 placeable parts.++## Signal chain++```+bridge -> [RF 2.49k 1%] -+- [CDIF 470n] -+- ADS1220 AINx+                         +- [CCM 47n 5%] -- GND (shared via per pair)+```++R sits at the contact end, C at the ADC end. A shared GND via per pair removes CM-to-DM conversion+error from CCM mismatch; `CDIF >= 10 x CCM` swamps the remainder.++## Filter constants++```+VIN(MAX)  = +/-[(AVDD - AVSS) - 0.4] / Gain  =  +/-36 mV at gain 128+                                                PGA output stays 200 mV off each rail+CM window = 1.25 - 4.16 V                       Eq 12/13/14, Eq 14 dominates+f_cm / f_diff = 2(CDIF/CCM) + 1 = 21            f_diff 64.8 Hz, f_cm 1360 Hz+```++Noise 0.41 µVpp at gain 128 / 20 SPS. −3 dB bandwidth 13.1 Hz at 20 SPS, 39.6 Hz at 90 SPS+(Table 8-4). Reference input current ±10 nA, buffered.++## Ratiometric topology++REFP0 (pin 7), AVDD (pin 10), MC16 and MC23 all sit on the same `+5V` node, downstream of F1.++- Fuse drop and the AP2112's ~70 mA are common to reference and excitation, so they cancel+- Resistance from that node out to the cell is outside the loop and appears as gain error —+  ~0.029 % for 0.1 Ω, calibratable++## Power++`+5V_IN` → **F1 polyfuse → +5V → D1 SMF6.0A → GND**. The fuse is upstream of the TVS, so a reverse+polarity fault drives TVS current through the fuse.++AP2112K-3.3 with 10 µF in (C5) and out (C6). C1 100 n sits 1.67 mm from AVDD.++## Indicator++**WS2812B-2020-V6** (`C52917434`). VIH is 0.55 × VDD and it runs from 3.3 V. Specifies 100 nF.+Not stocked at Mouser or DigiKey in the PLCC4 2020 package — LCSC only.++## Board state++0 DRC violations, 0 unconnected. All 30 contacts on the 2 mm lattice, MP datum 16.000 × 32.000 mm.++Three silk DRC checks are ignored. Five BOM/schematic LCSC mismatches exist in+`jlcpcb/project.db` — the BOM generates from that database, not the schematic fields.

@@ -1,4 +1,4 @@-<meta charset="utf-8">+<style>.overview-doc{font-size:14px;line-height:1.6}.overview-doc table{border-collapse:collapse;margin:10px 0}.overview-doc th,.overview-doc td{border:1px solid rgba(128,140,160,.3);padding:4px 10px;text-align:left}.overview-doc pre{background:rgba(128,140,160,.12);padding:10px;border-radius:6px;overflow-x:auto;font-size:12.5px}.overview-doc code{font-family:ui-monospace,monospace}</style><meta charset="utf-8"> <title>Feeder Reader — Feeder Pin Input Module</title> <style>   :root{ --bg:#151b23; --card:#1b232d; --ink:#e8eef5; --dim:#9fb0c3; --dim2:#7f92a6;@@ -622,6 +622,47 @@ addEventListener('resize', () => scene.getEngine().resize());     </div>   </div> +  +  <h2>Project overview</h2>+  <p class="prov">From the project's own overview document (author-written).</p>+  <div class="overview-doc">+<h1>RP2350A + ADS1220 Molecule — Function</h1>+<p>24-bit thrust acquisition. RP2350A (QFN-60) + ADS1220 + W25Q16JV + AP2112K-3.3, two differential+load-cell channels. 26.40 × 36.00 mm, 4-layer, 70 placeable parts.</p>+<h2>Signal chain</h2>+<pre><code>bridge -&gt; [RF 2.49k 1%] -+- [CDIF 470n] -+- ADS1220 AINx+                         +- [CCM 47n 5%] -- GND (shared via per pair)+</code></pre>+<p>R sits at the contact end, C at the ADC end. A shared GND via per pair removes CM-to-DM conversion+error from CCM mismatch; <code>CDIF &gt;= 10 x CCM</code> swamps the remainder.</p>+<h2>Filter constants</h2>+<pre><code>VIN(MAX)  = +/-[(AVDD - AVSS) - 0.4] / Gain  =  +/-36 mV at gain 128+                                                PGA output stays 200 mV off each rail+CM window = 1.25 - 4.16 V                       Eq 12/13/14, Eq 14 dominates+f_cm / f_diff = 2(CDIF/CCM) + 1 = 21            f_diff 64.8 Hz, f_cm 1360 Hz+</code></pre>+<p>Noise 0.41 µVpp at gain 128 / 20 SPS. −3 dB bandwidth 13.1 Hz at 20 SPS, 39.6 Hz at 90 SPS+(Table 8-4). Reference input current ±10 nA, buffered.</p>+<h2>Ratiometric topology</h2>+<p>REFP0 (pin 7), AVDD (pin 10), MC16 and MC23 all sit on the same <code>+5V</code> node, downstream of F1.</p>+<ul>+<li>Fuse drop and the AP2112's ~70 mA are common to reference and excitation, so they cancel</li>+<li>Resistance from that node out to the cell is outside the loop and appears as gain error —+  ~0.029 % for 0.1 Ω, calibratable</li>+</ul>+<h2>Power</h2>+<p><code>+5V_IN</code> → <strong>F1 polyfuse → +5V → D1 SMF6.0A → GND</strong>. The fuse is upstream of the TVS, so a reverse+polarity fault drives TVS current through the fuse.</p>+<p>AP2112K-3.3 with 10 µF in (C5) and out (C6). C1 100 n sits 1.67 mm from AVDD.</p>+<h2>Indicator</h2>+<p><strong>WS2812B-2020-V6</strong> (<code>C52917434</code>). VIH is 0.55 × VDD and it runs from 3.3 V. Specifies 100 nF.+Not stocked at Mouser or DigiKey in the PLCC4 2020 package — LCSC only.</p>+<h2>Board state</h2>+<p>0 DRC violations, 0 unconnected. All 30 contacts on the 2 mm lattice, MP datum 16.000 × 32.000 mm.</p>+<p>Three silk DRC checks are ignored. Five BOM/schematic LCSC mismatches exist in+<code>jlcpcb/project.db</code> — the BOM generates from that database, not the schematic fields.</p>+  </div>+   <h2>What it does</h2>   <p>A self-contained 24-bit acquisition molecule that <b>nests on the drone motor carrier</b> over its   machine-pin datum. It takes <span class="mono">+5&nbsp;V</span> in, makes its own

⋯ 1 unchanged line ⋯